I've been doing this a few years. Every time the subject of bypass capacitors comes up, someone always=20 repeats the gospel: "You need a variety of values, from several uF down=20 to hundreds of pF" or some such arrangement to get optimal bypassing. I don't agree. I think you [often] need a bulk capacitor whose size depends on where=20 your power supply is coming from. This is an unrelated subject in itself. But as far as localized capacitors go, I would like to posit that you=20 *only* need 100nF or 1uF capacitors. Unless you are doing RF bypassing. To back up this assertion I have measured several 0603 size capacitors=20 on a vector network analyzer. I saw nearly perfectly ideal capacitor=20 behavior from 10 kHz through 300 MHz. Now, in the microwave region you=20 have other much bigger issues to contend with so I haven't bothered to=20 exceed 300 MHz. The 1uF capacitor is for all measurable purposes the=20 same as the 1nF capacitor in the HF region up to the 300 MHz region.=20 There was no drop off at high frequencies for the 1uF cap. My testing shows that the ESL of these normal every-day capacitors is=20 incredibly low. The series inductance of the PCB traces matters MUCH=20 more, by at least a couple orders of magnitude. So I would conclude (I guess I should say "I posit") that to get the=20 best possible bypassing arrangement you merely need to make the PCB=20 trace inductance in the power supply loop (VDD at cap --> IC --> ground=20 at cap) as small as possible. The bulk capacitance available on the=20 board should be capable of handling the lower frequency (<10kHz) AC load. I will continue to liberally use 100nF and 1uF bypass capacitors and=20 throw in the odd 10nF cap for the 1.5 GHz PLL IC. Thinking any harder=20 about this will just give you a headache. I'd be happy to supply my VNA charts if anyone wants to see them. - Martin Klingensmith --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .